A generalization of the Birthday problem and the chromatic polynomial

Abstract

The birthday paradox states that there is at least a 50% chance that some two out of twenty-three randomly chosen people will share the same birth date. The calculation for this problem assumes that all birth dates are equally likely. We consider the following two modifications of this question. If the distribution of birthdays is non-uniform, does that increase or decrease the probability of matching birth dates? Further, what if we focus on birthdays shared by some particular pairs rather than any two people. Does a non-uniform distribution on birth dates increase or decrease the probability of a matching pair? In this paper we present our results in this generalized setting. We use some results and methods due to Sokal [17] concerning bounds on the roots of chromatic polynomials to prove our results.

Universal limit theorems in graph coloring problems with connections to extremal combinatorics

“…Define P G (p) to be the probability that G is properly colored. P G (p) is related to Stanley's generalized chromatic polynomial [50], and under the uniform coloring distribution it is precisely the proportion of proper c-colorings of G. Recently, Fadnavis [26] proved that P G (p) is Schur-convex for every fixed c, whenever the graph G is claw-free, that is, G has no induced K 1,3 . This implies that for claw-free graphs, the probability that it is properly colored is maximized under the uniform distribution, that is, p a = 1/c for all a ∈ [c].…”

Four Variations on Graded Posets

“…(3 + 1)-Avoiding Graded Posets (3+1)-avoidance is ubiquitous in the study of poset-avoidance, including the Stanley-Stembridge conjecture[SS93], the birthday problem[Fad11], etc. When we started…”